You can try any model you want, however I really suggest using <a target="_blank" rel="ugc" href="https://civarchive.com/models/260267/animagine-xl-v31">AnimagineXL 3.1 checkpoint </a>for XL and <a target="_blank" rel="ugc" href="https://civarchive.com/models/288584?modelVersionId=324524">AutismMix checkpoint </a>for Pony.Base model settings:<ul><li>Steps: &lt;30</li><li>Width: 896px</li><li>Height: 1132px</li><li>Sampler: Euler a</li><li>VAE: None</li><li>CFG Scale: 5 - 7</li><li>Clip skip: 2</li></ul>Upscaler settings:<ul><li>Upscaler: Latent (Nearest-Exact)</li><li>Upscale by: 1.5x</li><li>Denoising strength: 0.55</li></ul>LoRA settings:<ul><li>Weight: 0.7 - 1.0</li></ul>

Brammashell SDXL LoRA + Pony // GrandChase Dimensional Chaser

Really tried to make this work, but I'm struggling with low samples and bad quality samples for training... It works, sometimes. Anyway I think it's fine for now, might re-train in a future, for now I think it's good enough.

Files with SHA256: 6ee923198ad8b03c51b1199e2627631dda4042454b22041003920abe1b05e3f2